Any twenty-first century visit to a garden centre is about much more than plants. On a recent visit to one of these I paused just inside the shop, to I point out a cartoon image of a dog dressed as a human, cooking breakfast. Then I spotted another with a vacuum cleaner. Once I'd spotted one, they seemed to be everywhere. My younger companions quickly identified these cutesy images as AI generated, and dismissed them in a moment, decrying the taste and discrimination of both sellers and potential buyers. It's hard to disagree with them, but it's now also hard to ignore AI generated material.
I am already a casual user of generative AI to help in administrative tasks. I have no affiliation to or paid subscription for to any product, but I am exploring what the potential is for them. In search of some guidance I enlisted a chat bot but was shocked by the resulting line of questioning, when it asked whether I would be using the product mostly for creative writing? I knew, for example, that Hollywood script-writers were threatened enough by these tools to withdraw their labour in search for an agreement over banning these tools. Further, the controversy surrounding the 2026 Commonwealth Short Story Prize suggests many believe generative AI tools already write with literary merit. If not, the heated debate that followed about the author's own contribution is meaningless. So why were the architects of generative AI openly promoting such subversive abilities?
In search of more informed insight, I signed up for an online workshop on AI and creative writing. I expected some practical guidance, or at least some ethical discussion of these tools in creative practice. No such luck. What followed was a workshop, to a fairly standard model, one that played heavily on the significance of each individual's perspective and experience in informing the creative writing output. Fair enough, but the session remained silent on my the topics I'd hoped to explore.
That silence sadly does not remove the potential for Large Language Models and Generative AI tools to contribute to the creative process, or perhaps even their capacity to undertake such work alone. The architects of these tools are often demonised, with some justification. Suspect content is also often dismissed as AI Slop, again in many instances with some justification. But labour saving tech is tempting, so many human users are likely requesting output already, and much of it with use cases well beyond household administration.
However, some will still hold that any use of generative AI at any stage in creative work is unacceptable. Perhaps all with such views are by nature are Luddite, but the use of word processing software, capable of much more than merely recording the letters typed suggests otherwise. So does the existence and wide use of online dictionary and thesaurus. But there is a clear difference between such reference and checking tools and those that can generate page-ready text.
A further perhaps a more commonly held view that something is lost whenever generative AI is used to inform the literary endeavour. This might be driven by concern about the questionable ethics of sourcing LLM content, or over delegation of intellectual activity by the writer. Does this mean that generating ideas should be off limits? What about producing lists of potential near-rhyming words, or perhaps alternatives for the rhythm of a line? The case for considering the legitimacy and implications of using generative tools is pressing.
I don't have a fixed view on any of these questions yet, and have limited experience from which to make a contribution. It seems likely that writers, or readers, will disagree on this topic. But debate remains urgent for everyone who creates.